基于加权最小二乘谱迭积算法探测超导重力观测弱信号

摘要

提出采用加权最小二乘谱迭积算法探测超导重力观测数据隐藏的弱周期信号,该算法基于不等权的观测序列和序列的自相关性和互相关性,以高分辨率提取序列潜在的共同信号,削弱或抵消噪声信号.并以实例进行了分析和验证.%Least Squares Spectral Analysis ( LSSA) has been shown more suitable than Fourier analysis for analyzing time series with spikes, gaps, datum shrifts ( offsets) and other disturbances in a variety of applications in geodesy and geophysics. To detect the weak superconducting gravimetric signals, a methodology of product spectrum based on weighted LSSA is presented, which can take consideration the correlations among the signals and highlight the potential spectra power of high resolution in the common figures as well as weaken or eliminate the power of noise. A data set of superconducting gravimeter observations at the Canada Superconducting Gravimeter Installation (CSGI) has been assembled for analysis to show its efficacy.